A Note on Light Geometric Graphs

Total Page:16

File Type:pdf, Size:1020Kb

A Note on Light Geometric Graphs A note on light geometric graphs Eyal Ackerman∗ Jacob Foxy Rom Pinchasiz March 19, 2012 Abstract Let G be a geometric graph on n vertices in general position in the plane. We say that G is k-light if no edge e of G has the property that each of the two open half-planes bounded by the line through e contains more than k edges of G. We extend the previous result in [1] and withp a shorter argument show that every k-light geometric graph on n vertices has at most O(n k) edges. This bound is best possible. Keywords: Geometric graphs, k-near bipartite. 1 Introduction Let G be an n-vertex geometric graph. That is, G is a graph drawn in the plane such that its vertices are distinct points and its edges are straight-line segments connecting corresponding vertices. It is usually assumed, as we will assume in this paper, that the set of vertices of G is in general position in the sense that no three of them lie on a line. A typical question in geometric graph theory asks for the maximum number of edges that a geometric graph on n vertices can have assuming a forbidden configuration in that graph. This is a popular area of study extending classical extremal graph theory, utilizing diverse tools from both geometry and combinatorics. For example, an old result of Hopf and Pannwitz [3] and independently Sutherland [7] states that any geometric graph on n vertices with no pair of disjoint edges has at most n edges. This is a special case of Conway's thrackle conjecture. Let e be an edge of G. We say that G has a k-light side with respect to e, if one of the two open half-planes bounded by the line through e contains at most k edges of G. If G has a k-light side with respect to every edge e, then we say that G is k-light. In other words, G is k-light if no edge of G has the property that each of the two open half-planes bounded by the line through e contains more than k edges of G. The notion of a k-light graph is a weakening of the notion of a k-near bipartite graph defined in [1]. A graph G is k-near bipartite if every line in the plane bounds an open half plane containing at most k edges of G. Therefore, every k-near bipartite graphp is also a k-light graph. It is shown in [1] that k-near bipartite graphs on n vertices contain O( kn) edges. In this paper we prove the same result for k-light graphs, thus strengthening the result in [1]. Moreover, our proof is much shorter but on the other hand relies on other results about geometric graphs. ∗Department of Mathematics, Physics, and Computer Science, University of Haifa at Oranim, Tivon 36006, Israel. [email protected]. yDepartment of Mathematics, MIT, Cambridge, MA 02139-4307. Research supported by a Simons Fellowship and NSF grant DMS-1069197. [email protected]. zMathematics Department, Technion|Israel Institute of Technology, Haifa 32000, Israel. [email protected]. Supported by BSF grant (grant No. 2008290). 1 2 The maximum number of edges in k-light geometric graphs We are interested in the maximum numberp of edges of an n-vertex k-light geometric graph. n 2 A simple construction from [1] shows an n k lower bound for k ≤ 2 − 1 , even for k-near bipartite graphs. In this construction every line contains at most k edges of G in onep of the two open half-planes bounded by it. Another construction of a k-light graph with n k edges is obtained by taking thep vertices of a regular n-gon and connecting by edges vertices whose cyclic distance is at most k. In this construction, however, it is no longer true that every line bounds an open half-plane containing at most k edges of G. Our main result shows that these constructions are essentially best possible. Theoremp 1. Let n and k be positive integers. Every n-vertex k-light geometric graph has at most O(n k) edges. Proof. Let G be an n-vertex k-light geometric graph with m edges. We orient every edge e of G in such a way that the open half-plane bounded to the left of e contains at most k edges of G. Because G is k-light such an orientation exists. We will need the following two lemmas. Lemma 2.1. Let G be an oriented geometric graph on n vertices. There exists an absolute constant c3 such that if G has more than c3n edges, then it contains an edge e such that the open half-plane bounded to the left of e contains an edge of G. Proof. It is enough to show that in any (unoriented) geometric graph G with n vertices and sufficiently many (that is, at least c3n) edges there is an edge e such that each of the two open half-planes bounded by the line through e contains an edge of G. This is in fact the case k = 1 in Theorem 1 that we wish to prove. The reader is encouraged to find a simple proof of this fact. Here we will rely on a rather elaborate argument of Valtr [8] that proves a much stronger statement than what we need. We refer the reader to [5, 4, 8]. Two edges of a geometric graph are called avoiding or sometimes parallel if no line passing through one edge meets the other edge. Equivalently, two edges are avoiding if they are opposite edges in a convex quadrilateral. The notion of avoiding edges was first defined by Kupitz [5], who conjectured that any geometric graph on n vertices with more than 2n − 2 edges must contain a pair of avoiding edges. In [4] it is shown that if a graph G on n vertices does not contain a pair of avoiding edges, then the number of edges in G is at most 2n − 1. In [8] Valtr improved this bound by one, completing the proof of Kupitz' conjecture. He further generalized this result, showing that for any fixed k, every geometric graph with more than ckn edges contains k pairwise avoiding edges. Here ck is an absolute constant that depends only on k. In fact, Valtr's result is a bit stronger. Looking into the proof in [8] reveals that he actually shows that a geometric graph with more than ckn edges contains k edges e1; : : : ; ek that are pairwise avoiding, but what is more important to our needs is that the line through ei separates e1; : : : ei−1 from ei+1; : : : ; ek. More specifically, Valtr defines three partial orders on a set of edges in G and any chain with respect to any of the partial orders is a collection of such edges. It is then shown that if the number of edges in G is large enough, then there exists a chain of length k in one of the partial orders. Thus, for the case k = 3 it follows that if G contains more than c3n edges, then there are three pairwise avoiding edges e; f; g such that the line through f separates e and g. This immediately implies Lemma 2.1, as in any orientation of f the half-plane bounded to the left of f will contain an edge of G. Lemma 2.2. Let G be an oriented geometric graph on n vertices with m edges. There exists a positive absolute constant d with the following property. If the number of edges in G is greater 3 2 than 2c3n (where c3 is the constant from Lemma 2.1), then G contains at least dm =n pairs of edges (e; f) such that the open half-plane bounded to the left of e contains f. 2 Proof. This is by now a quite standard consequence of the result in Lemma 2.1 and is carried out by a similar probabilistic technique used to derive a similar bound for the number of pairs of crossing edges in a geometric graph (see p. 55 in [6], also p. 45 in [2]). Denote by x(G) the number of pairs of edges (e; f) in G such that the open half-plane bounded to the left of e contains f. Pick every vertex of G independently with probability p, and denote by G0 = (V 0;E0) the subgraph of G that is induced by the chosen vertices. Clearly, 0 0 2 0 4 E[jV j] = pn, E[jE j] = p m, and E[x(G )] = p x(G). On the other hand, it follows from 0 0 0 0 Lemma 2.1 that x(G ) ≥ jE j − c3jV j, and this holds also for the expected values: E[x(G )] ≥ 0 0 E[jE j] − c3E[jV j]. Plugging in the expected values and setting p = 2c3n=m < 1 we get that 1 m3 x(G) ≥ 2 2 . 8c3 n Let c3 and d be the constants from Lemmas 2.1 and 2.2. Clearly we may assume that G 3 2 contains at least 2c3n edges or else we are done. By Lemma 2.2, G contains at least dm =n pairs (e; f) of edges such that the open half-plane bounded to the left of e contains f. However, by the choice of orientation of the edges in G, an edge e can belong to at mostp k such pairs (e; f).
Recommended publications
  • On the Archimedean Or Semiregular Polyhedra
    ON THE ARCHIMEDEAN OR SEMIREGULAR POLYHEDRA Mark B. Villarino Depto. de Matem´atica, Universidad de Costa Rica, 2060 San Jos´e, Costa Rica May 11, 2005 Abstract We prove that there are thirteen Archimedean/semiregular polyhedra by using Euler’s polyhedral formula. Contents 1 Introduction 2 1.1 RegularPolyhedra .............................. 2 1.2 Archimedean/semiregular polyhedra . ..... 2 2 Proof techniques 3 2.1 Euclid’s proof for regular polyhedra . ..... 3 2.2 Euler’s polyhedral formula for regular polyhedra . ......... 4 2.3 ProofsofArchimedes’theorem. .. 4 3 Three lemmas 5 3.1 Lemma1.................................... 5 3.2 Lemma2.................................... 6 3.3 Lemma3.................................... 7 4 Topological Proof of Archimedes’ theorem 8 arXiv:math/0505488v1 [math.GT] 24 May 2005 4.1 Case1: fivefacesmeetatavertex: r=5. .. 8 4.1.1 At least one face is a triangle: p1 =3................ 8 4.1.2 All faces have at least four sides: p1 > 4 .............. 9 4.2 Case2: fourfacesmeetatavertex: r=4 . .. 10 4.2.1 At least one face is a triangle: p1 =3................ 10 4.2.2 All faces have at least four sides: p1 > 4 .............. 11 4.3 Case3: threefacesmeetatavertes: r=3 . ... 11 4.3.1 At least one face is a triangle: p1 =3................ 11 4.3.2 All faces have at least four sides and one exactly four sides: p1 =4 6 p2 6 p3. 12 4.3.3 All faces have at least five sides and one exactly five sides: p1 =5 6 p2 6 p3 13 1 5 Summary of our results 13 6 Final remarks 14 1 Introduction 1.1 Regular Polyhedra A polyhedron may be intuitively conceived as a “solid figure” bounded by plane faces and straight line edges so arranged that every edge joins exactly two (no more, no less) vertices and is a common side of two faces.
    [Show full text]
  • Graph Theory
    1 Graph Theory “Begin at the beginning,” the King said, gravely, “and go on till you come to the end; then stop.” — Lewis Carroll, Alice in Wonderland The Pregolya River passes through a city once known as K¨onigsberg. In the 1700s seven bridges were situated across this river in a manner similar to what you see in Figure 1.1. The city’s residents enjoyed strolling on these bridges, but, as hard as they tried, no residentof the city was ever able to walk a route that crossed each of these bridges exactly once. The Swiss mathematician Leonhard Euler learned of this frustrating phenomenon, and in 1736 he wrote an article [98] about it. His work on the “K¨onigsberg Bridge Problem” is considered by many to be the beginning of the field of graph theory. FIGURE 1.1. The bridges in K¨onigsberg. J.M. Harris et al., Combinatorics and Graph Theory , DOI: 10.1007/978-0-387-79711-3 1, °c Springer Science+Business Media, LLC 2008 2 1. Graph Theory At first, the usefulness of Euler’s ideas and of “graph theory” itself was found only in solving puzzles and in analyzing games and other recreations. In the mid 1800s, however, people began to realize that graphs could be used to model many things that were of interest in society. For instance, the “Four Color Map Conjec- ture,” introduced by DeMorgan in 1852, was a famous problem that was seem- ingly unrelated to graph theory. The conjecture stated that four is the maximum number of colors required to color any map where bordering regions are colored differently.
    [Show full text]
  • Archimedean Solids
    University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln MAT Exam Expository Papers Math in the Middle Institute Partnership 7-2008 Archimedean Solids Anna Anderson University of Nebraska-Lincoln Follow this and additional works at: https://digitalcommons.unl.edu/mathmidexppap Part of the Science and Mathematics Education Commons Anderson, Anna, "Archimedean Solids" (2008). MAT Exam Expository Papers. 4. https://digitalcommons.unl.edu/mathmidexppap/4 This Article is brought to you for free and open access by the Math in the Middle Institute Partnership at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in MAT Exam Expository Papers by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. Archimedean Solids Anna Anderson In partial fulfillment of the requirements for the Master of Arts in Teaching with a Specialization in the Teaching of Middle Level Mathematics in the Department of Mathematics. Jim Lewis, Advisor July 2008 2 Archimedean Solids A polygon is a simple, closed, planar figure with sides formed by joining line segments, where each line segment intersects exactly two others. If all of the sides have the same length and all of the angles are congruent, the polygon is called regular. The sum of the angles of a regular polygon with n sides, where n is 3 or more, is 180° x (n – 2) degrees. If a regular polygon were connected with other regular polygons in three dimensional space, a polyhedron could be created. In geometry, a polyhedron is a three- dimensional solid which consists of a collection of polygons joined at their edges. The word polyhedron is derived from the Greek word poly (many) and the Indo-European term hedron (seat).
    [Show full text]
  • Representing Graphs by Polygons with Edge Contacts in 3D
    Representing Graphs by Polygons with Side Contacts in 3D∗ Elena Arseneva1, Linda Kleist2, Boris Klemz3, Maarten Löffler4, André Schulz5, Birgit Vogtenhuber6, and Alexander Wolff7 1 Saint Petersburg State University, Russia [email protected] 2 Technische Universität Braunschweig, Germany [email protected] 3 Freie Universität Berlin, Germany [email protected] 4 Utrecht University, the Netherlands [email protected] 5 FernUniversität in Hagen, Germany [email protected] 6 Technische Universität Graz, Austria [email protected] 7 Universität Würzburg, Germany orcid.org/0000-0001-5872-718X Abstract A graph has a side-contact representation with polygons if its vertices can be represented by interior-disjoint polygons such that two polygons share a common side if and only if the cor- responding vertices are adjacent. In this work we study representations in 3D. We show that every graph has a side-contact representation with polygons in 3D, while this is not the case if we additionally require that the polygons are convex: we show that every supergraph of K5 and every nonplanar 3-tree does not admit a representation with convex polygons. On the other hand, K4,4 admits such a representation, and so does every graph obtained from a complete graph by subdividing each edge once. Finally, we construct an unbounded family of graphs with average vertex degree 12 − o(1) that admit side-contact representations with convex polygons in 3D. Hence, such graphs can be considerably denser than planar graphs. 1 Introduction A graph has a contact representation if its vertices can be represented by interior-disjoint geometric objects1 such that two objects touch exactly if the corresponding vertices are adjacent.
    [Show full text]
  • Graph Mining: Social Network Analysis and Information Diffusion
    Graph Mining: Social network analysis and Information Diffusion Davide Mottin, Konstantina Lazaridou Hasso Plattner Institute Graph Mining course Winter Semester 2016 Lecture road Introduction to social networks Real word networks’ characteristics Information diffusion GRAPH MINING WS 2016 2 Paper presentations ▪ Link to form: https://goo.gl/ivRhKO ▪ Choose the date(s) you are available ▪ Choose three papers according to preference (there are three lists) ▪ Choose at least one paper for each course part ▪ Register to the mailing list: https://lists.hpi.uni-potsdam.de/listinfo/graphmining-ws1617 GRAPH MINING WS 2016 3 What is a social network? ▪Oxford dictionary A social network is a network of social interactions and personal relationships GRAPH MINING WS 2016 4 What is an online social network? ▪ Oxford dictionary An online social network (OSN) is a dedicated website or other application, which enables users to communicate with each other by posting information, comments, messages, images, etc. ▪ Computer science : A network that consists of a finite number of users (nodes) that interact with each other (edges) by sharing information (labels, signs, edge directions etc.) GRAPH MINING WS 2016 5 Definitions ▪Vertex/Node : a user of the social network (Twitter user, an author in a co-authorship network, an actor etc.) ▪Edge/Link/Tie : the connection between two vertices referring to their relationship or interaction (friend-of relationship, re-tweeting activity, etc.) Graph G=(V,E) symbols meaning V users E connections deg(u) node degree:
    [Show full text]
  • Analytical Framework for Recurrence-Network Analysis of Time Series
    Analytical framework for recurrence-network analysis of time series Jonathan F. Donges,1, 2, ∗ Jobst Heitzig,1 Reik V. Donner,1 and J¨urgenKurths1, 2, 3 1Potsdam Institute for Climate Impact Research, P.O. Box 601203, 14412 Potsdam, Germany 2Department of Physics, Humboldt University Berlin, Newtonstr. 15, 12489 Berlin, Germany 3Institute for Complex Systems and Mathematical Biology, University of Aberdeen, Aberdeen AB24 3FX, United Kingdom (Dated: August 13, 2018) Recurrence networks are a powerful nonlinear tool for time series analysis of complex dynamical systems. While there are already many successful applications ranging from medicine to paleoclima- tology, a solid theoretical foundation of the method has still been missing so far. Here, we interpret an "-recurrence network as a discrete subnetwork of a \continuous" graph with uncountably many vertices and edges corresponding to the system's attractor. This step allows us to show that various statistical measures commonly used in complex network analysis can be seen as discrete estimators of newly defined continuous measures of certain complex geometric properties of the attractor on the scale given by ". In particular, we introduce local measures such as the "-clustering coefficient, mesoscopic measures such as "-motif density, path-based measures such as "-betweennesses, and global measures such as "-efficiency. This new analytical basis for the so far heuristically motivated network measures also provides an objective criterion for the choice of " via a percolation threshold, and it shows that estimation can be improved by so-called node splitting invariant versions of the measures. We finally illustrate the framework for a number of archetypical chaotic attractors such as those of the Bernoulli and logistic maps, periodic and two-dimensional quasi-periodic motions, and for hyperballs and hypercubes, by deriving analytical expressions for the novel measures and com- paring them with data from numerical experiments.
    [Show full text]
  • 1 Bipartite Matching and Vertex Covers
    ORF 523 Lecture 6 Princeton University Instructor: A.A. Ahmadi Scribe: G. Hall Any typos should be emailed to a a [email protected]. In this lecture, we will cover an application of LP strong duality to combinatorial optimiza- tion: • Bipartite matching • Vertex covers • K¨onig'stheorem • Totally unimodular matrices and integral polytopes. 1 Bipartite matching and vertex covers Recall that a bipartite graph G = (V; E) is a graph whose vertices can be divided into two disjoint sets such that every edge connects one node in one set to a node in the other. Definition 1 (Matching, vertex cover). A matching is a disjoint subset of edges, i.e., a subset of edges that do not share a common vertex. A vertex cover is a subset of the nodes that together touch all the edges. (a) An example of a bipartite (b) An example of a matching (c) An example of a vertex graph (dotted lines) cover (grey nodes) Figure 1: Bipartite graphs, matchings, and vertex covers 1 Lemma 1. The cardinality of any matching is less than or equal to the cardinality of any vertex cover. This is easy to see: consider any matching. Any vertex cover must have nodes that at least touch the edges in the matching. Moreover, a single node can at most cover one edge in the matching because the edges are disjoint. As it will become clear shortly, this property can also be seen as an immediate consequence of weak duality in linear programming. Theorem 1 (K¨onig). If G is bipartite, the cardinality of the maximum matching is equal to the cardinality of the minimum vertex cover.
    [Show full text]
  • An Enduring Error
    Version June 5, 2008 Branko Grünbaum: An enduring error 1. Introduction. Mathematical truths are immutable, but mathematicians do make errors, especially when carrying out non-trivial enumerations. Some of the errors are "innocent" –– plain mis- takes that get corrected as soon as an independent enumeration is carried out. For example, Daublebsky [14] in 1895 found that there are precisely 228 types of configurations (123), that is, collections of 12 lines and 12 points, each incident with three of the others. In fact, as found by Gropp [19] in 1990, the correct number is 229. Another example is provided by the enumeration of the uniform tilings of the 3-dimensional space by Andreini [1] in 1905; he claimed that there are precisely 25 types. However, as shown [20] in 1994, the correct num- ber is 28. Andreini listed some tilings that should not have been included, and missed several others –– but again, these are simple errors easily corrected. Much more insidious are errors that arise by replacing enumeration of one kind of ob- ject by enumeration of some other objects –– only to disregard the logical and mathematical distinctions between the two enumerations. It is surprising how errors of this type escape detection for a long time, even though there is frequent mention of the results. One example is provided by the enumeration of 4-dimensional simple polytopes with 8 facets, by Brückner [7] in 1909. He replaces this enumeration by that of 3-dimensional "diagrams" that he inter- preted as Schlegel diagrams of convex 4-polytopes, and claimed that the enumeration of these objects is equivalent to that of the polytopes.
    [Show full text]
  • Geometric Biplane Graphs I: Maximal Graphs ∗
    Geometric Biplane Graphs I: Maximal Graphs ∗ Alfredo Garc´ıa1 Ferran Hurtado2 Matias Korman3;4 In^esMatos5 Maria Saumell6 Rodrigo I. Silveira5;2 Javier Tejel1 Csaba D. T´oth7 Abstract We study biplane graphs drawn on a finite planar point set S in general position. This is the family of geometric graphs whose vertex set is S and can be decomposed into two plane graphs. We show that two maximal biplane graphs|in the sense that no edge can be added while staying biplane|may differ in the number of edges, and we provide an efficient algorithm for adding edges to a biplane graph to make it maximal. We also study extremal properties of maximal biplane graphs such as the maximum number of edges and the largest maximum connectivity over n-element point sets. 1 Introduction In a geometric graph G = (V; E), the vertices are distinct points in the plane in general position, and the edges are straight line segments between pairs of vertices. A plane graph is a geometric graph in which no two edges cross. Every (abstract) graph has a realization as a geometric graph (by simply mapping the vertices into distinct points in the plane, no three of which are collinear), and every planar graph can be realized as a plane graph by F´ary'stheorem [13]. The number of n-vertex labeled planar graphs is at least 27:22n n! [15]. · However, there are only 2O(n) plane graphs on any given set of n points in the plane [2, 22]. We consider a generalization of plane graphs.
    [Show full text]
  • A Multigraph Approach to Social Network Analysis
    1 Introduction Network data involving relational structure representing interactions between actors are commonly represented by graphs where the actors are referred to as vertices or nodes, and the relations are referred to as edges or ties connecting pairs of actors. Research on social networks is a well established branch of study and many issues concerning social network analysis can be found in Wasserman and Faust (1994), Carrington et al. (2005), Butts (2008), Frank (2009), Kolaczyk (2009), Scott and Carrington (2011), Snijders (2011), and Robins (2013). A common approach to social network analysis is to only consider binary relations, i.e. edges between pairs of vertices are either present or not. These simple graphs only consider one type of relation and exclude the possibility for self relations where a vertex is both the sender and receiver of an edge (also called edge loops or just shortly loops). In contrast, a complex graph is defined according to Wasserman and Faust (1994): If a graph contains loops and/or any pairs of nodes is adjacent via more than one line the graph is complex. [p. 146] In practice, simple graphs can be derived from complex graphs by collapsing the multiple edges into single ones and removing the loops. However, this approach discards information inherent in the original network. In order to use all available network information, we must allow for multiple relations and the possibility for loops. This leads us to the study of multigraphs which has not been treated as extensively as simple graphs in the literature. As an example, consider a network with vertices representing different branches of an organ- isation.
    [Show full text]
  • 5 Graph Theory
    last edited March 21, 2016 5 Graph Theory Graph theory – the mathematical study of how collections of points can be con- nected – is used today to study problems in economics, physics, chemistry, soci- ology, linguistics, epidemiology, communication, and countless other fields. As complex networks play fundamental roles in financial markets, national security, the spread of disease, and other national and global issues, there is tremendous work being done in this very beautiful and evolving subject. The first several sections covered number systems, sets, cardinality, rational and irrational numbers, prime and composites, and several other topics. We also started to learn about the role that definitions play in mathematics, and we have begun to see how mathematicians prove statements called theorems – we’ve even proven some ourselves. At this point we turn our attention to a beautiful topic in modern mathematics called graph theory. Although this area was first introduced in the 18th century, it did not mature into a field of its own until the last fifty or sixty years. Over that time, it has blossomed into one of the most exciting, and practical, areas of mathematical research. Many readers will first associate the word ‘graph’ with the graph of a func- tion, such as that drawn in Figure 4. Although the word graph is commonly Figure 4: The graph of a function y = f(x). used in mathematics in this sense, it is also has a second, unrelated, meaning. Before providing a rigorous definition that we will use later, we begin with a very rough description and some examples.
    [Show full text]
  • Penrose M. Random Geometric Graphs (OUP, 2003)(ISBN 0198506260)(345S) Mac .Pdf
    OXFORD STUDIES IN PROBABILITYManaging Editor L. C. G. ROGERS Editorial Board P. BAXENDALE P. GREENWOOD F. P. KELLY J.-F. LE GALL E. PARDOUX D. WILLIAMS OXFORD STUDIES IN PROBABILITY 1. F. B. Knight: Foundations of the prediction process 2. A. D. Barbour, L. Holst, and S. Janson: Poisson approximation 3. J. F. C. Kingman: Poisson processes 4. V. V. Petrov: Limit theorems of probability theory 5. M. Penrose: Random geometric graphs Random Geometric Graphs MATHEW PENROSE University of Bath Great Clarendon Street, Oxford OX26DP Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Auckland Bangkok Buenos Aires Cape Town Chennai Dar es Salaam Delhi Hong Kong Istanbul Kaarachi Kolkata Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi São Paulo Shanghai Taipei Tokyo Toronto Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries Published in the United States by Oxford University Press Inc., New York © Mathew Penrose, 2003 The moral rights of the author have been asserted Database right Oxford University Press (maker) First published 2003 Reprinted 2004 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organisation. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above.
    [Show full text]